JP5907342B2 - Vibrating transducer and method for manufacturing the vibrating transducer - Google Patents

Description

  The present invention relates to a vibratory transducer used for a transmitter or the like and a method for manufacturing the transducer.

  The vibration type transducer measures the applied physical quantity by detecting a change in the resonance frequency of the vibrator formed on the silicon substrate. Vibrating transducers are widely used in transmitters and the like as MEMS (Micro Electro Mechanical Systems) devices such as pressure sensors, acceleration sensors, angular velocity sensors, and resonators.

  FIG. 17 is a cross-sectional view illustrating the structure of a conventional vibration transducer. This figure is an example when the vibration type transducer is applied to a vibration type pressure sensor. As shown in this figure, in the vibration type pressure sensor 300, a vibrator 320 is formed on a silicon substrate 310 on which a diaphragm 311 is formed, and is covered with a shell 330. The inside of the shell 330 is a vacuum chamber 340 having a high degree of vacuum in order to increase the mechanical Q value of the vibrator 320 and is sealed by a sealing member 350 made of an oxide film.

  In the conventional vibration pressure sensor 300, the vibrator 320 and the vacuum chamber 340 surrounding the vibrator 320 are formed of a multilayer film. 18 and 19 show a manufacturing process of the conventional vibration type pressure sensor 300. FIG.

First, as shown in FIG. 18A, a silicon oxide film 401 is formed on an N-type silicon single crystal substrate 400 and patterned. Then, the portion from which the silicon oxide film 401 has been removed is undercut to form a recess, and selective epitaxial growth is performed with P-type silicon having a boron concentration of 10 ¹⁸ cm ⁻³ to grow P + single crystal silicon 402. Next, a P ++ single crystal silicon layer 403 is grown further upward on the surface of the P + single crystal silicon 402 by covering the surface of the P + single crystal silicon 402 with P-type silicon having a boron concentration of 3 × 10 ¹⁹ cm ⁻³ . Later, the P + single crystal silicon layer 402 becomes the gap below the vibrator 320, and the P ++ single crystal silicon layer 403 becomes the vibrator 320.

  Next, as shown in FIG. 18B, a silicon oxide film 404 is formed on the substrate surface including the P ++ single crystal silicon layer 403 and patterned. A portion indicated by a recess 405 from which the silicon oxide film 404 has been removed serves as a grounding portion of the shell 330 to the silicon substrate 310.

  As shown in FIG. 18C, a silicon nitride film 406 is formed on the substrate surface including the recess 405 and patterned. The silicon oxide film 404 and the silicon nitride film 406 on the P ++ single crystal silicon layer 403 form a gap on the vibrator 320.

  Next, as shown in FIG. 19A, P ++ polysilicon 407 is formed on the entire surface, and an etching solution introduction hole 408 for sacrificial layer etching is formed by patterning. This P ++ polysilicon 407 will later become a wiring for extracting the shell 330 and electrodes. The wiring can be formed by using P ++ / P + single crystal silicon or by diffusing impurities into the silicon substrate 400 before selective epitaxial growth.

  As shown in FIG. 19B, hydrofluoric acid is introduced from the etching solution introduction hole 408 to remove the silicon nitride film 406 and the silicon oxide film 404. Since the etching rate of the silicon nitride film 406 is low at the grounding portion of the shell 330 to the substrate, the silicon nitride film 406 serves as an etching stop layer in the lateral direction.

  Next, as shown in FIG. 19C, the P + single crystal silicon layer 402 is removed with an alkaline solution (hydrazine, KOH, TMAH, etc.). At this time, the P ++ single crystal silicon layer 403 and the P ++ polysilicon 407 are not etched because impurities are introduced at a high concentration. Thereby, the vibrator 320 is formed.

As shown in FIG. 19D, a sealing member 350 (for example, SiO ₂ or glass formed by sputtering) is formed by sputtering, vapor deposition, CVD, epitaxial growth, or the like, and the etching solution introduction hole is blocked. A fine vacuum chamber 340 is formed.

  Then, as shown in FIG. 19E, P ++ polysilicon 407 is patterned to form a shell 330, and electrical wiring from the vibrator 320 and the shell 330 is formed, and an electrode for a bonding pad is formed. Thereafter, the silicon substrate 400 is thinned from the back surface and the diaphragm 311 is formed, whereby the vibration type pressure sensor 300 as shown in FIG. 17 is obtained.

JP 2005-37309 A

  As described above, the vibrator 320 of the conventional vibration transducer is formed through the formation of the sacrificial layer (P + single crystal silicon layer 402, silicon oxide film 404, silicon nitride film 406) and the sacrificial layer etching process. The In addition, in order to increase the mechanical Q value of the vibrator 320, the sealing member 350 for vacuum sealing is formed.

  For this reason, a sacrificial layer forming step, a sacrificial layer etching step, and a film forming step for vacuum sealing cannot be indispensable in the manufacturing process, and there is a great restriction on simplification of the steps.

  Accordingly, an object of the present invention is to simplify the manufacturing process of the vibrator of the vibratory transducer.

In order to solve the above-described problem, a vibratory transducer according to a first aspect of the present invention includes a second-type semiconductor layer in which a second-type semiconductor layer and a first-type semiconductor layer are sequentially stacked on a first-type semiconductor substrate. Further, the vibrator and the electrode are formed with a space therebetween, and a gap is provided on the first-type semiconductor substrate side and the first-type semiconductor layer side of the vibrator.
Here, the first-type semiconductor substrate, the second-type semiconductor layer, and the first-type semiconductor layer can be made of a material that does not include a single laminated member that constitutes the layer.
The second type semiconductor layer may be a high concentration impurity layer.
Further, an SOI (Silicon On Insulator) substrate may be used instead of the first type semiconductor substrate.
In order to solve the above problems, a method of manufacturing a vibration type transducer according to a second aspect of the present invention includes a film forming step of sequentially stacking a second type semiconductor layer and a first type semiconductor layer on a first type semiconductor substrate, An internal region including the second-type semiconductor layer, including an etching step of forming two trenches having a wide upper portion and a lower portion in parallel and an annealing step of performing an annealing process in a hydrogen atmosphere. It is characterized by.
Here, in the etching step, anisotropic etching and isotropic etching can be repeatedly performed.

  According to the present invention, since a sacrificial layer forming step, a sacrificial layer etching step, and a film forming step for vacuum sealing are not required, the manufacturing process of the vibrator of the vibratory transducer can be simplified.

It is a figure explaining the structure of the vibration type pressure sensor of this embodiment. It is a top view of a vibration type pressure sensor. It is sectional drawing which shows the structure of the pad for vibrator | oscillators, and the pad for electrodes. It is sectional drawing which shows the structure of the pad for guards. It is sectional drawing which shows the structure of the pad for N board | substrates. It is sectional drawing which shows the structure of the pad for upper N layers. It is a figure explaining the 1st-3rd process of the manufacturing method of the vibration type pressure sensor concerning this embodiment. It is a figure explaining the 4th-6th process of the manufacturing method of the vibration type pressure sensor concerning this embodiment. It is a figure explaining the 7th-9th process of the manufacturing method of the vibration type pressure sensor concerning this embodiment. It is a figure explaining an etching process. It is a figure explaining an annealing process. It is a figure which shows the example of a mask pattern. It is a figure explaining the usage condition of a vibration type pressure sensor. It is a figure explaining another example of an etching process. It is a figure explaining the etching process which raised the precision. It is a figure which shows the case where a vibrator | oscillator is formed on an SOI substrate. It is sectional drawing explaining the structure of the conventional vibration type transducer. It is a figure explaining the manufacturing process of the conventional vibration type pressure sensor. It is a figure explaining the manufacturing process of the conventional vibration type pressure sensor.

  Embodiments of the present invention will be described with reference to the drawings. In the present embodiment, an example in which a vibration transducer is applied to a vibration pressure sensor will be described. FIG. 1 is a view for explaining the structure of the vibration type pressure sensor of this embodiment, and FIG. 1A is a cross-sectional view of the vibration type pressure sensor 100. However, in order to make the structure easy to understand, the dimensional ratio is different from the actual one.

  As shown in the figure, the vibration pressure sensor 100 has a four-layer structure in which a P ++ layer 120, an upper N layer 130, and an oxide film 140 are stacked on an N substrate 110. Each layer on the N substrate 110 is a single layer generated in one film formation process. Of course, a structure in which an N ++ layer, an upper P layer, and an oxide film are stacked on a P substrate may be employed.

  A diaphragm 111 that is bent by pressure input is formed on the N substrate 110. A vibrator 121a is disposed on the P ++ layer 120 on the diaphragm 111, and a pressure that is input by measuring a resonance frequency that changes according to the pressure is measured by a drive detection circuit (not shown). However, it can also be configured as an absolute pressure sensor without the diaphragm 111.

  1B is a plan view of the P ++ layer 120 of the vibration pressure sensor 100 as viewed from above, and FIG. 1C is a cross-sectional view as viewed from a plane orthogonal to the cross-sectional view of FIG. It is. 1A is a cross-sectional view taken along line A in the plan view shown in FIG. 1B, and FIG. 1C is a cross-sectional view taken along line B in the plan view shown in FIG. Further, the plan view shown in FIG. 1 (b) is a plane obtained by cutting the P ++ layer 120 along the line H in the sectional view of FIG. 1 (a) and the line V in the sectional view of FIG. 1 (c). Represents. In FIG. 1, white portions indicate cavities generated by the etching process, and a portion other than a through hole 125 for exposing an N substrate pad described later is in a vacuum state.

  This example has a structure in which two vibrators are arranged, which are referred to as a first vibrator 121a and a second vibrator 121b, respectively, and collectively referred to as a vibrator 121 when they are not distinguished from each other. The other parts provided in plural are also referred to in the same manner. Moreover, although this embodiment has exemplified a capacitive vibration transducer, the present invention can also be applied to a vibration transducer using a permanent magnet.

  The vibrator 121 formed on the P ++ layer 120 in the surface layer portion of the vibration pressure sensor 100 is surrounded by a vacuum chamber 126, and electrodes 123 are arranged on both sides of the vacuum chamber 126. Specifically, the first vibrator 121a is sandwiched between the electrode 123a and the electrode 123b, and the second vibrator 121b is sandwiched between the electrode 123c and the electrode 123d. Each electrode 123 is provided with an electrode pad 153 made of aluminum or the like, and the electrode pad 153 is exposed on the upper surface.

  The first vibrator 121a is connected to the first vibrator electrode 122a, and the second vibrator 121b is connected to the second vibrator electrode 122b. The first vibrator electrode 122a is provided with a first vibrator pad 152a, the second vibrator electrode 122b is provided with a second vibrator pad 152b, and both vibrator pads 152 are exposed on the upper surface. Yes.

  Both the vibrator 121 and the electrode 123 that need to be electrically separated are formed in the P ++ layer 120, but both are insulated by a cavity gap 127. Each electrode 123 and the vibrator electrode 122 are also insulated from the guard 124 (the ground portion of the P ++ layer 120) and the gap 127. A vacuum chamber 126 is formed by a gap 127 around the vibrator 121.

  As shown in this figure, the vibration type pressure sensor 100 of the present embodiment has a three-layer structure of an N substrate 110, a P ++ layer 120, and an upper N layer 130, and each layer is a single layer constituting a layer. It is comprised with the material which does not contain other than a lamination member. This is because another material such as a sealing member is not necessary for the upper layer portion, and another material such as an insulating material for ensuring insulation is not necessary for the middle layer portion. However, you may make it comprise each layer including another material.

  FIG. 2 is a top view of the vibration pressure sensor 100. The upper surface of the vibration type pressure sensor 100 is covered with an oxide film 140, and vibrator pads 152a, 152b, electrode pads 153a, 153b, 153c, 153d, and a card pad 154, each formed of aluminum or the like. The N substrate pad 155 and the upper N layer pad 156 are exposed.

  The vibrator pad 152 is a pad for wiring to the vibrator 121, and the electrode pad 153 is a pad for wiring to the electrode 123. Both pads can have the same structure. FIG. 3 is a cross-sectional view showing the structure of the vibrator pad 152 and the electrode pad 153. As shown in this figure, the vibrator pad 152 and the electrode pad 153 are attached to the vibrator electrode 122 and the electrode 123 formed in the P ++ layer 120 at the place where the upper N layer 130 and the oxide film 140 are removed. It is done.

  The guard pad 154 is a pad for wiring the guard 124. FIG. 4 is a cross-sectional view showing the structure of the guard pad 154. As shown in the figure, the guard pad 154 is attached to the guard 124 formed on the P ++ layer 120 at the place where the upper N layer 130 and the oxide film 140 are removed.

  The N substrate pad 155 is a pad for wiring the N substrate 110. FIG. 5 is a cross-sectional view showing the structure of the N substrate pad 155. As shown in the figure, the N substrate pad 155 is attached to the N substrate 110 at the place where the upper N layer 130, the P ++ layer 120, and the oxide film 140 are removed.

  Upper N layer pad 156 is a pad for wiring to upper N layer 130. FIG. 6 is a cross-sectional view showing the structure of the upper N layer pad 156. As shown in the figure, the upper N layer pad 156 is attached to the upper N layer 130 at the place where the oxide film 140 is removed.

  Next, a method for manufacturing the vibration pressure sensor 100 according to the present embodiment will be described. Here, referring to FIGS. 7 to 9, the manufacturing method is divided into nine steps for each of the vibrator 121, the vibrator pad 152 and the electrode pad 153, the upper N layer pad 156, and the N substrate pad 155. Separately described.

  First, the first to third steps will be described with reference to FIG. These steps are common to the vibrator 121, the vibrator pad 152 and the electrode pad 153, the upper N layer pad 156, and the N substrate pad 155. In the first step, an N-type silicon substrate is prepared. The N-type silicon substrate becomes the N substrate 110 through a later process.

  In the second step, a P ++ single crystal layer and an N single crystal layer are epitaxially grown on the surface of the N-type silicon substrate. Through the subsequent steps, the P ++ single crystal layer becomes the P ++ layer 120, and the N single crystal layer becomes the upper N layer 130. In the third step, the surface is covered with an oxide film, photolithography is performed, and then the oxide film is etched. At this time, patterning is performed so as to obtain a shape as shown in FIG.

  Next, the fourth to sixth steps will be described with reference to FIG. In the fourth step, etching is performed by DRIE (deep reactive ion etching) using the oxide film patterned in the third step as a mask. In the patterning in the third step, the vibrator 121 is spaced so that a desired gap can be obtained in the subsequent fourth and fifth steps, and the vibrator pad 152 and the electrode pad 153 are provided in the electrode portion and the wiring. The interval is such that the necessary width can be obtained as a part. The upper N layer pad 156 is masked so as not to be etched in this step, and the N substrate pad 155 is patterned so that the entire surface is etched to form an opening.

  As shown in FIG. 8, in the etching of the fourth step, the portion near the surface layer of the oxide film is etched narrowly, and the vicinity of the upper and lower boundaries of the P ++ layer is etched wider by increasing the lateral etching amount.

  Therefore, for example, as shown in FIG. 10A, the portion of the upper N layer 130 close to the surface layer of the oxide film is subjected to directional etching so as to become a narrow groove 161, and at the boundary portion with the P ++ layer 120, Isotropic etching is performed so as to form a wide groove 162 as shown in FIG. Further, after forming a medium groove 163 as shown in FIG. 10C by directional etching, a wide groove 164 is formed at the boundary with the N substrate 110 as shown in FIG. 10D. Isotropic etching is performed.

  In the fifth step, the surface oxide film is removed, and in the sixth step, annealing is performed in a hydrogen atmosphere. By annealing in a hydrogen atmosphere, surface silicon atoms move so as to minimize the surface energy.

  For this reason, as shown in FIG. 11A, the vibrator 121 has a narrow interval between the two parallel grooves 165a and 165b, so that the wide groove portions are joined to each other. Further, the width of the middle groove portion connecting the wide groove portions is widened and smoothed. Furthermore, the narrow groove portion near the surface layer is closed. As a result, as shown in FIG. 11B, the vibrator 121 and the surrounding space are formed, and a gap is generated. In addition, by performing annealing in a hydrogen atmosphere, the space around the vibrator 121 is vacuum-sealed to become a vacuum chamber 126. Thereby, vibration with a high mechanical Q value can be realized.

  Further, in the vibrator pad 152 and the electrode pad 153, as shown in FIG. 11C, the gap between the two grooves 165c and 165d is wide, so that the wide groove portions are not joined to each other, and the wide grooves are not joined. The width of the middle groove part connecting the parts widens and becomes smooth. In addition, a narrow groove near the surface layer is closed. As a result, an independent space is formed as shown in FIG. The N substrate pad 155 has a smooth wide open wall.

  Next, the seventh to ninth steps will be described with reference to FIG. In the seventh step, the surface is oxidized, and the oxide film in the vibrator pad 152, the electrode pad 153 region, and the guard pad 154 region (not shown) is removed by photolithography and oxide film etching. Then, using the oxide film as a mask, anisotropic etching is performed using an alkali solution capable of etching with a concentration difference such as hydrazine or EDP. As a result, the P ++ layer 120 is exposed. Note that RIE may be performed instead of anisotropic etching using an alkaline solution.

  In the eighth step, the oxide film used for the mask in the seventh step is removed, and a surface oxide film is formed again. This layer becomes the oxide film 140. In the ninth step, the oxide film is patterned to form a contact hole, aluminum is deposited, and the aluminum in the contact portion is patterned.

  As described above, the vibration type pressure sensor 100 according to the present embodiment does not require a sacrificial layer forming step, a sacrificial layer etching step, and a film forming step for vacuum sealing. Can be simplified. Conventionally, about five masks are required until the vacuum sealing film forming process, and the vibration pressure sensor 100 according to the present embodiment performs the third process and the fourth process before the annealing process in the sixth process. A single mask used in the above is sufficient. In this respect, the manufacturing process can be greatly reduced. Furthermore, management items can be reduced by reducing the number of manufacturing processes, which can contribute to an improvement in product yield.

  FIG. 12 shows a mask pattern example of one mask used in the third step and the fourth step. As shown in the figure, the mask pattern can be drawn with intermittent lines. In the example of this figure, the A-part enlarged view shows two patterns of broken lines and dotted lines as examples of intermittent lines. However, if the opening is closed at the time of annealing and the part that needs to be separated is separated, The shape may also be

FIG. 13 is a diagram for explaining how the vibration pressure sensor 100 is used. As shown in FIG. 13A, in order to electrically insulate the upper N layer 130 and the N substrate 110 of the vibration pressure sensor 100 from the P ++ layer 120, it can be used in a circuit so as to be in a reverse bias condition. A maximum voltage Vh is applied to the upper N layer 130 and the N substrate 110. Bias voltages of + Vb and −Vb having the same magnitude are applied to the electrode 123 facing the vibrator 121. The vibrator 121 is pulled with equal force from both sides by electrostatic attraction generated by this bias voltage. On the other hand, an AC signal _vac having a frequency equal to the resonance frequency is input to the vibrator 121. At this time, the potential between the electrode 123 and the vibrator 121 changes, and a force expressed by [Equation 1] works between the vibrator 121 and the electrode 123.
As can be seen from [Equation 1], since a force proportional to the AC signal v _ac acts on the vibrator 121, the vibrator 121 vibrates at the same frequency as the AC signal v _ac . This vibration changes the electrostatic capacity between the vibrator 121 and the electrode 123 in order to displace the vibrator 121. By detecting this change in capacitance by an external circuit, the vibration of the vibrator 121 can be extracted as a change in voltage. In the example of FIG. 13A, this capacitance change is converted into a voltage by using a current-voltage conversion circuit, and the capacitance of the capacitance between the vibrator 121 and the electrode 123 is obtained by taking the differential of this signal. A voltage output proportional to the change is obtained.

In the example of FIG. 13A, the vibrator 121 is driven by the AC signal _vac having the same frequency as the resonance frequency. However, since the resonance frequency is not actually known, as shown in the block diagram of FIG. The vibrator 121 is vibrated using a self-excited circuit. The circuit A in FIG. 13A corresponds to the block A in FIG. By applying positive feedback to the block A by the amplifier β, the vibrator 121 vibrates by itself. At this time, when controlling the amplitude, the amplitude is adjusted using an excitation control circuit provided outside the positive feedback pool. The excitation control circuit controls the output amplitude of the block A in the operational amplifier B by comparing the reference voltage Vref with the amplitude.

  In the above embodiment, in the fourth step of the vibrator, as shown in FIG. 11A, etching is performed so that the wide groove portions of the grooves 165a and 165b are close to each other, and annealing in a hydrogen atmosphere is performed. The wide groove portions were joined to each other by the processing so that the vibrator 121 was separated, but as shown in FIG. 14A, etching was performed so that the wide groove portions of the grooves 165a and 165b were connected, The vibrator 121 may be separated before annealing in a hydrogen atmosphere. Also in this case, as shown in FIG. 14B, the vibrator 121 and the vacuum chamber 126 are formed by annealing in a hydrogen atmosphere.

  Further, in the above embodiment, as shown in FIG. 10, by repeating anisotropic etching and isotropic etching, a narrow groove portion, a wide groove portion, a medium groove portion, and a wide groove portion are provided. However, a mask made of an oxide film may be used in combination in order to increase the system of the shape of the groove formed by DRIE. For example, as shown in FIG. 15A, after the upper N layer is etched vertically by anisotropic etching, thin oxidation is performed as shown in FIG. As shown, the bottom oxide film is removed by dry anisotropic etching. Then, as shown in FIG. 15D, the P ++ layer 120 is widely etched by DRIE. Thereafter, oxidation and dry anisotropic etching are repeated (FIGS. 15E to 15H), whereby a desired structure can be obtained with high accuracy.

In the above embodiment, each layer is formed on the N substrate 110. However, instead of the N substrate 110, each layer is formed on the SOI substrate using an SOI wafer (Silicon On Insulator). Also good. FIG. 16 shows a case where a vibrator is formed on an SOI substrate. In the example of this figure, as shown in FIG. 16A, an SOI substrate 170 made of N or P type silicon 171, a box layer SiO ₂ 172, and an active layer P type silicon 173 is used.

In this case, as shown in FIG. 16B, a P ++ epitaxial layer 181, an N-type epitaxial layer 182, and an oxide layer 183 are formed on the SOI substrate 170. When the vibrator portion is etched by the above-described procedure, an etch stop is generated at the portion where the P-type silicon 173 is in contact with the SiO ₂ 172 as shown in FIG. After the etch stop, the lateral etching is continued to narrow the width under the vibrator.

After that, when annealing is performed in a hydrogen atmosphere, etching is performed by sublimation of SiO _{2 at} the time of sealing, and silicon atoms move at a portion where stress concentration occurs at the interface with SiO _2. Since the occurrence of drooling occurs, the vibrator has a structure as shown in FIG. Even such a structure does not affect the characteristics of the vibrator.

  In addition, when it is necessary to lengthen the sealing time, the bottom of the vacuum chamber around the vibrator is further distorted as shown in FIG. Even if it exists, it is possible to ensure performance as a vibrator.

DESCRIPTION OF SYMBOLS 100 ... Vibration type pressure sensor, 110 ... N board | substrate, 111 ... Diaphragm, 120 ... P ++ layer, 121 ... Vibrator, 122 ... Vibrator electrode, 123 ... Electrode, 124 ... Guard, 125 ... Through-hole, 126 ... Vacuum chamber, 127: gap, 130: upper N layer, 140: oxide film, 152: vibrator pad, 153 ... electrode pad, 154 ... card pad, 155 ... N substrate pad, 156 ... upper N layer pad, 161 ... groove, 162 ... groove, 163 ... groove, 164 ... groove, 165 ... groove, 170 ... SOI substrate, 171 ... N / P silicon, 172 ... SiO _2, 173 ... P-type silicon, 181 ... P ++ epitaxial layer, 182 ... N-type epitaxial layer, 183 ... oxide layer, 300 ... vibration type pressure sensor, 310 ... silicon substrate, 311 ... diaphragm, 320 ... vibrator, 330 ... shi 340 ... vacuum chamber 350 ... sealing member 400 ... silicon substrate 400 ... N-type silicon single crystal substrate 401 ... silicon oxide film 402 ... P + single crystal silicon layer 403 ... P ++ single crystal silicon layer 404 ... Silicon oxide film, 405 ... concave, 406 ... silicon nitride film, 407 ... P ++ polysilicon, 408 ... etching liquid introduction hole

Claims (5)

The second type semiconductor layer and the first type semiconductor layer are sequentially stacked on the first type semiconductor substrate without sandwiching the insulating layer ,
A vibrator and an electrode are formed in the second type semiconductor layer with a space therebetween, and a gap is provided on the first type semiconductor substrate side and the first type semiconductor layer side of the vibrator. Vibrating transducer.   2. The first-type semiconductor substrate, the second-type semiconductor layer, and the first-type semiconductor layer are all made of a material that does not include a single laminated member that constitutes the layer. The vibratory transducer as described.   The vibratory transducer according to claim 1, wherein the second type semiconductor layer is a high concentration impurity layer.   A film forming step of sequentially stacking a second type semiconductor layer and a first type semiconductor layer on the first type semiconductor substrate;
  An etching step of forming, in parallel, two grooves having wide upper and lower portions in an internal region including the second type semiconductor layer;
  An annealing process for annealing in a hydrogen atmosphere;
A method of manufacturing a vibratory transducer characterized by comprising:   5. The method of manufacturing a vibration type transducer according to claim 4, wherein the etching step repeatedly performs anisotropic etching and isotropic etching.